The Kinematics Library was initiated and is currently developed at the Robotics Innovation Center of the German Research Center for Artificial Intelligence (DFKI) in Bremen.
A framework for handling serial robot kinematics. This framework contains the following kinematic solvers: kdl, trac_ik, ikfast, analytical solver for 7 dof manipulator and an optimisation-based kinematic solvers. The aim of this framework is to offer the user to use different types kinematic solvers by just changing the configuration.
This framework offers the following kinematics solvers to solve both forward and inverse kinematics.
- KDL : Use Orocos Kinematics and Dynamics library to solve the kinematic problem.
- TRACIK: Tracik offers a better and faster solution compared to KDL. This solver combines KDL and optimization-based solver to solve the kinematics problem.
- If you want to use trac_ik solver, please install the tracik library before building this library.
- IKFAST: ikfast is an analytical kinematic solver for non-redundant robots.
- The user needs to provide a "ikfast_generated.so" of the robot.
- 7DOF Analytical solvers: There are two different methods to analytically solve the inverse kinematic for a 7DOF arm. But the manipulator needs to follow a certain structure.
- IK7DOF: shimizu_method - SRS (Spherical - revolute - Spherical joints)
- SRS: asfour_method - SRS (Spherical - revolute - Spherical joints)
- OPT: Optimization-based solver
- Please install NLOPT in order to use this solver.
- The inverse kinematic problem is formualted as a wieghted sum optimization problem.
- Currently position, orientation, velocity, and acceleration erms are optimized. The user can able to set the weight for the objective funtions.
- The forward kinematics is solved using the KDL library.